2 hours ago, guenseli said:

I have contacted them, but got a not very satisfying answer:

"Hallo! Das ist die RenderingEngine des Simulators und hat mit FSGUNG nichts zu tun. Je naeher man kommt, desto genauer zeichnet der Simulator das Mesh. Je nach Fluggeschwindigkeit und Rechnerleistung ist das mehr oder weniger sichtbar. Da muessen Sie sich leider an die Simulatorhersteller wenden."

 

Translation:

"That is because of the rendering engine of the simulator. Depends on flying speed and computer power. You have to contact the developer of the simulation."

 

In my ears it sounds more like: doesn't interest me.

It really is a limitation of the engine. It will only draw max LOD for a limited radius, then pop in the lower LOD.

What I would recommend is a higher LOD_radius in the P3D.cfg. I use something like 8.5 and I find it's a good balance between performance and fidelity.

Thank you for that insight, Daniel, you are right. This is the first time ever I read someone to clearly point to the root of the issue.

I made an analysis by changing nearly all parameters,including even tesselation in Prepar3d3 without success - texture morphing always was there, maybe a bit less or more, but always distracting.

Now LOD was not an option to choose from at that time, thus I missed it. Just to confirm your theory, I set LOD to 11 in Prepar3d4 and mesh resolution to 10 m, and pudding morphing is completely gone. This would be one more reason to change to Prepar3d4, if I didn't already,

These are certainly questionable parameters, but they clearly show the root of the issue.

Kind regards, Michael

3 hours ago, pmb said:

I set LOD to 11 in Prepar3d4 and mesh resolution to 10 m, and pudding morphing is completely gone

Considering that not allowing an LOD>6.5 was only a limitation in P3d V3 and not in either P3d V4, FSX or FSX-SE, is the solution that simple? If so, then the problem is no more.

4 hours ago, pmb said:

Now LOD was not an option to choose from at that time, thus I missed it. Just to confirm your theory, I set LOD to 11 in Prepar3d4 and mesh resolution to 10 m, and pudding morphing is completely gone. This would be one more reason to change to Prepar3d4, if I didn't already,

Thank you Michael...I want to give this a try too. I have a question though...in prepar3d.cfg file, the setting MESH_RESOLUTION=23. If I do not want to change this value via the normal graphics menu in prepar3dv4, but change via notepad++, would I change this to MESH_RESOLUTION=10 ??

Or better, what is your mesh resolution value set to in your prepar3d.cfg file?

Hi,

my usual setting is MESH_RESOLUTION=22 corresponding to an in-sim setting of 5m which ORBX suggests. I only set it to 10m for the test above. Of course you can set mesh-resolution arbitrarily low, in which case you will certainly get rid of all morphing issues -  together with all mesh structures, unfortunately. It's certainly a tradeoff.

I no way do I insist in these to be ideal or even suggested values. I only wanted to confirm there's a new parameter available now which does matter in this issue - as Daniel observed before.

Kind regards, Michael

 

5 hours ago, pmb said:

Thank you for that insight, Daniel, you are right. This is the first time ever I read someone to clearly point to the root of the issue.

I made an analysis by changing nearly all parameters,including even tesselation in Prepar3d3 without success - texture morphing always was there, maybe a bit less or more, but always distracting.

Now LOD was not an option to choose from at that time, thus I missed it. Just to confirm your theory, I set LOD to 11 in Prepar3d4 and mesh resolution to 10 m, and pudding morphing is completely gone. This would be one more reason to change to Prepar3d4, if I didn't already,

These are certainly questionable parameters, but they clearly show the root of the issue.

Kind regards, Michael

Are you talking about the LOD_RADIUS?

I've been experimenting with it, at a setting of 3.5 P3D is using 6 gigs of system ram and at 6.5 P3D is using 9 gigs of system ram. I've gone as far as 30 and P3D never goes past that 9 gigs of system ram. I think P3Dv4 still restricts lod to a max of 6.5. Hopefully they change that.

I'm using FsGlobal Ultimate NG mesh at 1M and I find my mesh popping to be very good (almost completely gone) I think the mesh popping comes down to how much our systems are being taxed at the time.

I've been able to significantly improve it but I don't believe it's been by a single silver bullet. It's been by balancing my performance. Around the PNW I've got my main CPU thread with an AM of 245 averaging around 60-70% and my GPU about the same. I haven't totally eliminated mesh popping but you really have to be looking for it.

Maybe Rob A, will weigh in here. I too use FSGUNG and I don't see any significant mesh popping, but now I'm curious as to whether the LOD does go above 6.5 in P3d V4. I agree with you, that the "mesh popping" issue is far more likely to be related to short term events when the hardware is heavily stressed.

11 minutes ago, jabloomf1230 said:

Maybe Rob A, will weigh in here. I too use FSGUNG and I don't see any significant mesh popping, but now I'm curious as to whether the LOD does go above 6.5 in P3d V4. I agree with you, that the "mesh popping" issue is far more likely to be related to short term events when the hardware is heavily stressed.

I would really like to make good use of some of the 64gigs of ram I bought ;)

6 hours ago, laserit said:

Are you talking about the LOD_RADIUS?

Indeed.

Basically your experience is just the opposite, as I see it. My test flight is from KSFO to KHAF which is within FTX NCA. I know the places where morphing potentially happens on this short hop over the hills.

I didn't much more testing so far. I own Pilot's Ultimate, too, but have it deactivated within ORBX regions.

Maybe others can confirm/deny what happens. Not sure if system load is involved, but who knows. When I made the tests in Prepar3d once, I dialed back a lot of sliders without any success (this was another machine than in my sig now).

If memory serves me right, we didn't observe this "feature" in FSX nor in Prepar3d1. I supposed it has been introduced with Prepar3d2 in conjunction with tessalation. Now the simple answer would be switching off tesselation altogether. I tested this in Prepar3d3, too, and while it may have improved things a bit it didn't completely eliminate them. Plus scenery (notably water) without tesselation looks even worse than with some morphing.

Kind regards, Michael

4 hours ago, pmb said:

Indeed.

Basically your experience is just the opposite, as I see it. My test flight is from KSFO to KHAF which is within FTX NCA. I know the places where morphing potentially happens on this short hop over the hills.

I didn't much more testing so far. I own Pilot's Ultimate, too, but have it deactivated within ORBX regions.

Maybe others can confirm/deny what happens. Not sure if system load is involved, but who knows. When I made the tests in Prepar3d once, I dialed back a lot of sliders without any success (this was another machine than in my sig now).

If memory serves me right, we didn't observe this "feature" in FSX nor in Prepar3d1. I supposed it has been introduced with Prepar3d2 in conjunction with tessalation. Now the simple answer would be switching off tesselation altogether. I tested this in Prepar3d3, too, and while it may have improved things a bit it didn't completely eliminate them. Plus scenery (notably water) without tesselation looks even worse than with some morphing.

Kind regards, Michael

If my memory serves me right, mesh popping has been an irritating issue for as long as I've been using 3rd party high definition mesh. I can't remember which I used first, I had the lifetime membership for FSGenesis and I was ecstatic when Holger Sandmann released his British Columbia meshes.

One thing is for certain, an LOD of 6.5 uses the exact same amount of system ram as 9.5, 11, 15 and 30. When using an LOD of 6.5 or less, the amount of system ram will change accordingly.

Sincerely

Floyd Stolle

There's two types of terrain popping reported: Mountains appearing when we are approaching, and small patches of ground morphing below the plane. The mountains appear due to coming into range of the area around the user where number of points increase, a mechanism tied to LOD. There is a finite amount of resource to handle the increasing number of mesh points - too many points simply brings that pop-up place nearer the user. We could reduce resolution of the input mesh rather than draw radius, it is tied in so changing radius alters the outcome as has been found.

1 hour ago, SteveW said:

There's two types of terrain popping reported: Mountains appearing when we are approaching, and small patches of ground morphing below the plane. The mountains appear due to coming into range of the area around the user where number of points increase, a mechanism tied to LOD. There is a finite amount of resource to handle the increasing number of mesh points - too many points simply brings that pop-up place nearer the user. We could reduce resolution of the input mesh rather than draw radius, it is tied in so changing radius alters the outcome as has been found.

Steve youve found a new technical area to try to explain simply to us idiots, now that you've done affinity masks to death (and very well I might add!) :p

Cheers K

A three year old necro post?

 

nope